First public alpha release of TNH Scholar, providing a suite of AI-powered text processing tools.
Core Features
TNH-FAB Command Line Tool
- Text punctuation and formatting with AI assistance
- Intelligent section analysis and breakdown
- Line-by-line translation with context preservation
- Pattern-based text processing with Git-backed version control
- Support for processing by sections or paragraphs
Audio Processing Tools
- YouTube audio download and processing
- Audio chunk splitting with silence or sentence boundary detection
- Transcription pipeline with OpenAI Whisper
- Support for batch processing of multiple files
Utility Tools
- `ytt-fetch`: YouTube transcript fetching utility
- `nfmt`: Newline formatting tool
- `token-count`: OpenAI token counting utility
- `tnh-setup`: Project configuration setup tool
Installation
bash
pip install tnh-scholar
After installation, run setup:
bash
tnh-setup
**Note:** Requires Python 3.12.4 and OpenAI API credentials
Optional Features
Additional functionality available through optional dependencies:
- OCR capabilities: `pip install "tnh-scholar[ocr]"`
- GUI tools: `pip install "tnh-scholar[gui]"`
- Query features: `pip install "tnh-scholar[query]"`
- Development tools: `pip install "tnh-scholar[dev]"`
Documentation
Documentation (working-in-progress) available at: https://aaronksolomon.github.io/tnh-scholar/
Requirements
- Python 3.12.4
- OpenAI API key (for AI-powered features)
- FFmpeg (for audio-transcribe processing)
Known Limitations
- Incomplete and possibly inaccurate documentation
- Currently supports Python 3.12.4 only
- Audio-transcribe requires additional system dependencies: FFmpeg for audio
- Type checking still in development (see mypy_errors.txt)
Acknowledgments
Special thanks to:
- Thay Phap Luu and Steve Suellenthorpfor inspiration for the project
- Daniel Miessler's 'fabric' project for pattern system inspiration